Data through: Oct 31, 2026 · updated Nov 5 · Updates: Weekly · Next: Nov 12
WThe three bars show the last 3 weeks for this indicator, oldest left to newest right. Bar height reflects each reading relative to the other two — the tallest is the highest of the three, the shortest the lowest.
↓ 4.5% MoM↓ 17.1% YoY
Historical context: Currently in the 1st percentile — near historically low/favorable levels.
What this means right now: Initial jobless claims are very low — below 220K. Claims have been this low in only about 1 week in 10 since 1967 (326 of 3,122 weeks), mostly in the late 1960s and since 2017 (117 weeks in 1967–69, 3 in 1973, 206 since 2017). Only one of those weeks came during a recession.
Initial Jobless Claims · Weekly · 1967–2026
NBER recessions
Scroll to read · zoom and pan in Expand
Monthly values
Initial Jobless Claims — the last 12 months
Every month's reading, with the band the heatmap gives it. The chart above shows the same series in full.
Month
Value
Band
Oct 2026
189K
Very strong reading
Sep 2026
199K
Very strong reading
Aug 2026
207K
Very strong reading
Jul 2026
198K
Very strong reading
Jun 2026
217K
Very strong reading
May 2026
225K
Healthy reading
Month-end readings of the same series as the chart. Values are the current record, not point-in-time: a month that has since been revised shows its revised value.
What is the Jobless Claims?
Initial jobless claims count the number of workers who filed for unemployment benefits for the first time in the prior week. Released every Thursday morning for the week ended the Saturday before — five days after the reference week — it is one of the site's two weekly labor market series, with continuing claims a week behind it.
When claims are low (below 250,000 a week), few workers are losing their jobs and the job market is healthy. When claims jump, layoffs are picking up. The four-week average smooths out the weekly noise. Seven of the eight recessions from 1967 to 2020 pushed claims to 400K or above. A four-week average that stayed between 325K and 400K for a month or more has been a weaker warning. Since 1967 it has been followed by higher unemployment within six months in 6 of 22 cases through September 2026, each on the edge of a recession.
How We Color-Code the Jobless Claims
Our heatmap colors this indicator against historically significant thresholds:
Below 220K
Very low — below 220K a week; minimal layoffs
220K – 270K
Low — 220K to 270K; layoffs contained
270K – 325K
Moderate — 270K to 325K; a moderate level of labor market churn
325K – 400K
Elevated — 325K to 400K; layoffs rising
400K or above
High — 400K or above. Seven of the eight recessions of 1967–2020 pushed claims this high (1970's worst week was 374K), but more than half the times the four-week average crossed it, no recession came within a year (12 of 21 to September 2026)
Historical Extremes — What Happened Next?
When this indicator reaches extreme levels, history shows consistent patterns:
Apr 2020COVID Shutdown
6.14M
The most claims in a single week in the record since 1967 — 6.1 million in the week ending April 4, 2020, as COVID shutdowns spread (in the revised record). The week before was first reported at 6.65 million on April 2, 2020, and now reads 5.95 million. The previous record was 695K, in 1982.
Sep 2022Historic Low
190K
The fewest claims since September 1969 at the time (in the revised record) — almost no one was being laid off in the very tight post-COVID job market. The week ending April 2, 2022 was first reported at 166K; after revision it reads 214K.
Mar 2009Financial Crisis Peak
665K
Claims stayed at or above 600K in 18 of the 21 weeks from late January to June 2009. Layoffs on that scale, month after month, pushed unemployment to 10% over the following 12+ months.
Frequently Asked Questions
Why are jobless claims released weekly when most data is monthly?
The Department of Labor collects unemployment insurance filing data from all 50 states weekly as an administrative byproduct of the benefits system. Because states already collect this data to process benefit payments, the DOL can aggregate and publish it weekly with minimal additional effort. This administrative collection lets the DOL publish each week's count five days after the week ends.
What level of initial claims is "normal"?
It depends on the era. In the five years before COVID (2015–2019) a typical week was 245K, and nine weeks in ten fell between 210K and 280K. The early 2010s ran far higher as the post-crisis layoffs faded (a typical week was 291K over 2010–2019). Since 2022 the typical week has been 218K. Below 220K is the site's lowest band — 58 weeks in 2015–2019 and 136 since 2022. Weeks above 350K have been common over the whole record — about 2 weeks in 5 since 1967 (1,342 of 3,122), most of them outside recessions. That is why the site's elevated band starts at 325K and its high band at 400K. The four-week average smooths out holiday and weather effects.
Why do claims spike around holidays?
Several factors cause holiday-related distortions. Seasonal layoffs in retail and hospitality follow the major shopping seasons; state unemployment offices have backlogs around holidays; and seasonal adjustment factors can create anomalies. This is why analysts focus on the 4-week moving average and are cautious about interpreting single-week readings around Thanksgiving, Christmas, and Easter.
How quickly do rising claims predict rising unemployment?
Over the whole cycle, claims run about three months ahead of the unemployment rate: their yearly changes match best at that lead (correlation 0.83, against 0.72 in the same month). As a warning, though, they are weaker. When the four-week average stayed between 325K and 400K for a month or more (22 times from 1967 to September 2026), unemployment was 0.3 point higher within six months in only 6 — each on the edge of a recession. At the turns into recession the record is mixed. Ahead of the eight since 1969, claims (the four-week average, 20% above its year's low) rose before the unemployment rate had risen 0.3 point in three cases and after it in five.
Do all workers who are laid off file for initial claims?
No — not all laid-off workers file. Workers who receive substantial severance packages often delay filing. Workers who believe they'll quickly find another job may not bother. Self-employed workers and some contract workers are ineligible. As a result, initial claims undercount total layoffs — JOLTS layoffs data provides a more complete picture of total involuntary separations.